Return to DTVCH6 | Return to EUonline Home
Editor's note: All illustrations here will appear in a separate browser
window for your convenience, using Adobe's PDF format. File sizes are noted in small-font
italics: (000K).
You can install a PDF viewer for free by downloading it from http://www.adobe.com
This document is Copyright (C) 1999 by National Public Radio
On behalf of its member stations, with help from the Public Telecommunications Facilities Program, National Public Radio undertook one the most aggressive spectrum analysis studies carried out to date by public radio. On July 24, 1998 the final report was released. It is the purpose of this paper to review basic methodology, to summarize the results of the study and to discuss the impact.
Purpose and Scope:
The non-commercial educational FM band runs from 88.1 MHz to 91.9 MHz. Channel-six TV stations operate near the bottom of the NCE FM band, using a 6 MHz bandwidth from 82 to 88 MHz. There exists considerable historic experience of noncommercial educational FM stations causing interference to channel-six TV stations. Until the completion of this study there was no data available to predict the interference impact on NCE FM stations of the new channel-six digital television assignments. It was not within the scope of this study to determine if NCE FM stations would cause interference to DTV stations assigned to channel-six.
Unlike standard or NTSC television, DTV’s digital modulation technique makes full use of the six megahertz TV channel with high-energy modulation. If we were to look at the spectral energy within an NTSC TV channel, we would see three carrier peaks for video, audio and color information, the rest of the spectrum would appear relative underutilized. However, the view into the DTV channel spectrum presents an entirely different picture. The full six-megahertz bandwidth is charged with high-energy spikes literally from the beginning to the end of the channel. The challenge to DTV transmitter manufacturers and operators is to contain this highly charged signal to the assigned frequency bandwidth without spillage. If the spillage occurs and is on the high end of channel-six’s assigned frequency bandwidth our NCE FM stations will suffer. Determining just how deeply NCE FM stations would be affected was the purpose of the research.
Study techniques and methods:
| FM Receiver Measurements | The study can be divided into two parts,
laboratory measurement and "real world" predictions. The first part was done by
Tom Keller and his staff at the Digital Radio Test Laboratory. Keller, whose lab has been
involved in testing RDBS and digital radio, studied six high-quality brand name receivers,
the Audiovox, Delco, Denon RDS, Denon NAB, Panasonic and the Pioneer. Also studied were three sub-carrier receivers. These receivers were examined for their abilities to reject radio frequency emissions from channel-six DTV stations. Once the receivers’ rejection capabilities were measured, a ratio of the signal levels from a DTV station that would cause interference to a given NCE FM station was calculated. This ratio of a "desired" FM signal strength to an "undesired" DTV signal strength is called the D/U ratio. To put it in another way, the FM receivers were measured to determine what the D/U ratio would have to be in order to maintain a constant signal to noise ratio of 55 dB. At this level, the DTV noise can be plainly heard as a constant low level shushing sound in quiet passages under music and talk. The D/U characteristics of the receivers were totaled in an averaging method to develop the figures shown in table 1. This table graphs the sliding D/U ratio that just begins to cause audible noise above the 55 dB threshold at various frequencies from 88.1 to 90.3 MHz. Interference beyond 90.3 MHz was deemed inconsequential. Notice that at 88.1 MHz the undesired DTV signal strength would have to be 8 dB below the FM station’s signal in order to maintain the 55 dB signal to noise ratio. While at 90.3 MHz the unwanted DTV signal can be up to 33 dB higher than the desired FM station signal without exceeding the 55 dB signal to noise ratio. See Table 1 in PDF format of FCC Feb 23, ’98, Out-of-channel Mask. (26K) The Keller study also looked at the impact of DTV interference on FM multipath. Multipath occurs when two or more signals from the original station mix in a manner that is destructive to the original signal. This happens when a signal from a station reaches the receiving antenna from a direct path and one or more reflected paths. Path reflections are usually caused by hills, tall buildings, water towers and other similar objects. The multipath studies determined that channel-six DTV out-of-band splatter exacerbated the multipath effect on nearby FM channels from 88.1 to 88.9 MHz. In multipath situations, the increased interference effect of the multipath requires an increase in the D/U ratio in order to maintain the desired 55 dB signal to noise ratio. The tiny cross-hatched line on the D/U graph ,shown in table 1, represents the increase in D/U which results when the FM receiver operates within a multipath environment. A DTV transmitter emission mask dictates the amount of fall off of radio frequency energy of a DTV signal as it passes outside the 6 MHz channel to which a station is assigned. As the Keller study was being implemented, the FCC made a change in the DTV emission mask requirement that tightened up the amount of out-of-band emissions a DTV station was allowed. Although, there were five masks studied, the out-of-band emissions of three of the five were selected for the mapping phase of the study.
|
| Mapping the Interference | Once the receiver D/U characteristics were
measured and a D/U table was prepared, the firm of Hammett & Edison was hired to
prepare maps that showed the locations where several NCE FM stations were predicted to
receive interference from DTV stations in selected markets. The maps used the TIREM
propagation prediction model. TIREM stands for Terrain Integrated Rough Earth Model. It is
argued that this model improves on the better known Longley-Rice prediction method to
produce excellent results. The Hammett & Edison TIREM implementation used the USGS
3-second digital terrain elevation database. The method itself considers all significant
characteristics along a radio path to calculate the impact on a radio signal. A
three-meter FM receiving antenna height was assumed. Using the TIREM model to plot the
locations where the D/U ratio for a given station operating on a given channel at a known
power was exceeded, the mapping program produced a series of color coded interference
areas or "pools". Generally, more of these pools of interference are found when
the FM receiver is located closer to the DTV station. There are numerous factors that affect the amount of interference to an FM station. These include the ratio of power between the DTV station and the FM station. (The higher the DTV station’s power is in relationship to the FM station’s power the more interference will be caused.) The farther removed in frequency of the FM station from the low edge of the band, the smaller the resulting areas of interference. The smoother the terrain is or the better the path is between the DTV station and the listener the worse the interference will be. The worse the path is between the FM station and the listener the greater the interference is likely to be. So the TIREM maps reflect numerous physical and electrical factors which combine to affect the location and abundance of interference.
|
| The maps | Each of the maps presented in the study has
color-coded plots of interference which represent three main DTV emission masks, New FCC
standard (Post Feb ’98), a "New technology" mask modeled by Comark on its
DTV transmitters and an "Out of specifications" mask having the characteristics
of a malfunctioning transmitter. The DTV interference areas of three NCE FM stations in Connecticut and seven stations in the Philadelphia area were mapped. On each map three color-codes were used that represent the interference areas caused by the DTV station under each of the masks. The red color code was used for the "New technology" mask. This is a "best case" mask under which the DTV out-of-band emissions have been carefully controlled. Therefore, for the stations studied, the interference areas defined by the red color is very likely to occur regardless of the DTV emission mask employed. This interference will only occur when the DTV station’s undesired signal is quite large in comparison to the desired FM signal strength. Since the D/U ratio for perceptible interference improves when the FM station is further removed from the DTV station’s frequency, the values for the interference plotted in the red, yellow/orange and green areas will change from map to map depending on the FM station’s frequency. In general, the yellow/orange areas plot the locations of predicted interference using the FCC’s new standard DTV mask while the green areas mark the locations there interference would occur if a DTV transmitter were malfunctioning. The interference areas are upwardly inclusive, in other words, interference that would be caused under the "New FCC standard" mask would also include the interference in the red area caused under the "New technology" mask. Likewise, the interference that would occur under the "Out of specifications" mask would include both the red and the yellow areas as well. |
The following is a brief review of each of the Hammet & Edison maps:
A.) WNHU, 88.7 MHz, West Haven, CT. Power 1.7 kW ERP, 49 meters HAAT. WBNT-DTV6, 1 kW, 314 M HAAT. (136K)
This map shows that an area around the DTV transmitter and a small area to the north (Along West Rock Ridge) within WNHU’s 60 dBu (red color) will receive interference from WBNT-DT even when the superior "New technology" mask is employed. In addition to the red area, an area of interference using the "New FCC standard" mask is predicted in a low-lying zone near North Haven (Orange color.) This map, as does most of the others, shows numerous areas where interference will be caused if the DTV transmitter operates using the "Out of specifications" mask that simulates a malfunctioning transmitter. When using the "New technology" mask some 65 people will receive interference within the WNHU 60 dBu signal contour. This is a total of only .02 percent. However, a total of 9,834 people are predicted to receive interference when the "New FCC standard" mask is employed. This is 2.5 percent of the population within the station’s 60 dBu signal.
B.) WPKN, 89.5 MHz , Bridgeport, CT, Power 10 kW, 169 meters HAAT, WBNT-DTV6, 1 kW (125K)
This map shows a red and yellow area immediately around the DTV transmitter where interference to WPKN will be received. Larger areas of interference are predicted to be received under the "Out of specifications" mask. (Green.) Interference will be caused to 9,457 people even when the "New technology" mask is employed (Red). A total of 19,872 people will be caused interference under the "New FCC standard" mask (orange)
C.) WPKT, 90.5 MHz, Meriden, CT, Power 18.5 kW, 251 meters HAAT, WBNT-DTV, 1 kW (101K)
This map shows a small area of interference having 801 people if the channel-six DTV station is operating using the "New technology" emissions mask but a larger area under when the DTV station is operating using the "New FCC standard" mask. The population in the "New FCC standard" mask DTV interference area amounts to 279,529 people which is 15.9 percent of the total population available within the WPKT 60 dBu. Clearly, in this case, the implementation of an emissions mask stricter than the "New FCC standard" will prevent a lot of headache for WPKT.
D.) WPEB, 88.1 MHz, Philadelphia , PA, Power 1 watt, 60 meters HAAT, WPVI-DTV, 8.77 kW, 332 M HAAT (101K)
Few, if any other, public radio stations operate with such a small effective radiated power as one watt. This map shows that if WPVI-DTV implements the "New FCC standard" emissions mask, the interference area to WPEB will be over 96.2 percent of the FM station’s 60 dBu coverage. Use of the "New technology" emissions mask reduces the population receiving interference to 109,383 people which is 73.6 percent of the station’s 60 dBu population. Clearly, no matter how tight the emission mask is set, this station will suffer severe interference. It is WPEB’s combination of very low power and frequency as well as its close proximity to the DTV channel-six TV station that cause its dilemma.
E.) WXPN, 88.5 MHz, Philadelphia, PA, Power 5 kW, 280 meters HAAT, WPVI-DTV, 8.77 kW, 332 meters HAAT. (128K)
No interference is predicted to be caused WXPN if the DTV station adopts the "new technology " mask. However, Interference to 132,948 people, or 2.89 percent of the population within WXPN’s 60 dBu coverage area occurs when the DTV station employs the "New FCC standard" mask.
F.) WBYO, 88.7 MHz, Philadelphia, PA, Power .9 kW, 133 meters HAAT, WPVI-DTV, 8.77 kW, 332 meters HAAT (117K)
Use of the FCC standard DTV mask results in interference to 46,243 people, which amounts to 25.5 percent of the station’s 60 dBu coverage area population. If the "new technology" mask is used, the interference area is reduced to affecting 1,412 people which is .78 percent of the station’s 60 dBu population.
G.) WYBF, 89.1 MHz. Radnor Township, PA, Power .7 kW, 68 meters HAAT, WPVI-DTV, 8.77 kW, 332 meters HAAT (137K)
This map shows substantial interference will be caused under the "New FCC standard" mask and the "New technology" emissions mask. With the "New Technology" mask in place, interference is caused to 102,342 people within the WYBF 60 dBu. This amounts to 39.6 percent of the total. Using the "New FCC standard" mask interference is caused to 140,990 people or 54.5 percent of the total.
Using the "New Technology" mask a total of 4,954 people are predicted to receive interference. This is 1.15 percent of the population within the 60 dBu coverage area. The "New FCC standard" mask delivers interference to 85,647 people or 19.8 percent of the 60 dBu coverage.
I.) WRTI, 90.1 Mhz. Philadelphia, PA, Power 12.5 kW, 308 meters HAAT, WPVI-DTV, 8.77 kW, 332 meters HAAT (129K)
If the "New technology" or the "New FCC standard" emissions mask is employed, no interference is caused to WRTI by WPVI-DTV. The map indicates that interference to some 257,317 people will be caused to WRTI if the "Out of specifications" mask is employed.
J.) WHYY, 90.9 Mhz, Philadelphia, PA, Power 13.5 kW, 280 meters HAAT, WPVI-DTV, 8.77 kW, 332 meters HAAT (128K)
Since 90.9 MHz is so far removed from the low end of the FM band and WHYY operates with considerable power and antenna height, the maps show no interference when the either the "New Technology" mask or the "New FCC standard" mask is employed. The maps show that a small amount of interference is caused under the "New FCC standard" emission mask.
| What does this mean for my station? | The total impact on NCE FM is unknown at this
time because we don’t have a good count of the number of TV markets where stations
will transmit DTV on channel-six. We do know that there were only three new DTV assignments made where channel-six was assigned as the operating channel. The cities to which these assignments were made are Bridgeport, CT, Washington, DC and Toledo, OH. However, the problem may go significantly further than those three markets. The Commission has stated that it will allow all TV broadcasters to return to their original channel if they so desire. Since WPVI has been assigned channel 64 and that channel is between 52 and 69 and therefore outside the core, the station will have no choice other than to return to the use of its original NTSC channel six at the end of the FCC prescribed transition period. There are some 190 television stations that have been assigned DTV channels outside the core, twelve of these stations are channel-six NTSC facilities. The twelve stations to which we refer are: WCTV-DT, 52, Thomasville, GA, Complicating the matter are the 44 other NTSC, channel-six, TV stations licensed in the U.S. Any of these stations may give up its newly assigned DTV channel if it deems that it would rather transmit DTV on its original channel. So, it is possible that some of these channel-six NTSC stations may move DTV operations to channel-six at some future date causing more of the kind of interference exhibited during this study. Further, the Commission has stated that it may allow higher DTV power operation on channel-six for those stations that qualify. The propagation prediction studies show that NCE stations occupying frequencies at the low end of the band, operating relative low power in relationship to the DTV station and having antennas that are located in terrain sensitive areas will have serious problems with interference from channel-six DTV television stations. This problem will be more severe in populated areas were the transmitter for the channel-six DTV station is close at hand. The testing also showed that multipath situations, such as those found in populated urban areas around tall buildings or in mountainous areas exacerbated the interference.
|
| Some mitigating factors | Based on the results of this
study, the impact of the DTV is to degrade the signal to noise ratio of the affected FM
station. The Keller testing also determined that in the presence of DTV interference the
signal to noise ratio was improved significantly by transmitting in monaural. It should be noted that TIREM assumed horizontal polarization for both the DTV station and the FM station and many FM stations operate using elements of horizontal and vertical polarization. It is likely that the severity of the predicted interference will decline with mobile receiver reception since most automobile listening occurs in vehicles with radios that use vertically polarized antennas. The Comark "New Technology" mask promises to significantly improve a DTV transmitter’s ability to reject out-of-band emissions. Use of "digital adaptive pre-correction" to shape the emission mask through ongoing digital sampling is an excellent idea. Hopefully, the other transmitter manufacturers will find similar methods to provide highly controlled rejection of out-of-band emissions.
|
| Conclusion | The continued use of channel-six by TV stations for DTV transmissions poses a significant threat to many non-commercial educational FM stations throughout the country. |
Doug Vernier is Director, Broadcasting Services, University of Northern Iowa, Cedar Falls Iowa. He is a past member of the NPR Board where he served in a number of capacities including Chairman of the Program Committee. Doug is also President of V-Soft Communications a broadcast technical consulting firm with an emphasis in technical propagation software development and sales. Doug is also a ham operator, K0DV. Reach himl at vernier@uni.edu or dvernier@v-soft.com.